import torchfrom transformers import AutoTokenizer, AutoModelForCausalLMfrom peft import PeftModel base_model_name = "TinyLlama/TinyLlama-1.1B-Chat-v1.0"adapter_name = "Vipplav/tinyllama-finetuned-faq" # 1. Load tokenizer and base modeltokenizer = AutoTokenizer.from_pretrained(base_model_name)base_model = AutoModelForCausalLM.from_pretrained( base_model_name, torch_dtype=torch.float16 if torch.cuda.is_available() else torch.float32, device_map="auto") # 2. Attach Vipplav LoRA adaptermodel = PeftModel.from_pretrained(base_model, adapter_name)model.eval() # 3. Format FAQ Promptprompt = "<|system|>\nYou are a helpful enterprise support assistant. Answer the customer FAQ concisely.</s>\n<|user|>\nHow can I reset my account password?</s>\n<|assistant|>\n" inputs = tokenizer(prompt, return_tensors="pt").to(model.device)with torch.no_grad(): outputs = model.generate( **inputs, max_new_tokens=150, temperature=0.3, repetition_penalty=1.1, eos_token_id=tokenizer.eos_token_id ) response = tokenizer.decode(outputs[0][inputs.input_ids.shape[1]:], skip_special_tokens=True)print("Response:", response.strip())